-
NOAA Technical Memorandum NMFS
OCTOBER 2006
STEELHEAD OF THE SOUTH-CENTRAL/SOUTHERN CALIFORNIA COAST:
POPULATION CHARACTERIZATION FOR RECOVERY PLANNING
David A. Boughton Peter B. Adams Eric Anderson Craig Fusaro
Edward Keller Elise Kelley Leo Lentsch
Jennifer Nielsen Katie Perry
Helen Regan Jerry Smith Camm Swift
Lisa Thompson Fred Watson
NOAA-TM-NMFS-SWFSC-394
U.S. DEPARTMENT OF COMMERCE National Oceanic and Atmospheric
Administration National Marine Fisheries Service Southwest
Fisheries Science Center
-
NOAA Technical Memorandum NMFS
The National Oceanic and Atmospheric Administration (NOAA),
organized in 1970, has evolved into an agency which establishes
national policies and manages and conserves our oceanic, coastal,
and atmospheric resources. An organizational element within NOAA,
the Office of Fisheries is responsible for fisheries policy and the
direction of the National Marine Fisheries Service (NMFS).
In addition to its formal publications, the NMFS uses the NOAA
Technical
Memorandum series to issue informal scientific and technical
publications when complete formal review and editorial processing
are not appropriate or feasible. Documents within this series,
however, reflect sound professional work and may be referenced in
the formal scientific and technical literature.
-
NOAA Technical Memorandum NMFS This TM series is used for
documentation and timely communication of preliminary results,
interim reports, or special pur-pose information. The TMs have not
received complete formal review, editorial control, or detailed
editing.
OCTOBER 2006
STEELHEAD OF THE SOUTH-CENTRAL/SOUTHERN CALIFORNIA COAST:
POPULATION
CHARACTERIZATION FOR RECOVERY PLANNING
David A. Boughton1, Peter B. Adams1, Eric Anderson1, Craig
Fusaro2, Edward Keller3, Elise Kel-ley4, Leo Lentsch5, Jennifer
Nielsen6, Katie Perry7, Helen Regan8, Jerry Smith9, Camm
Swift10,
Lisa Thompson11, and Fred Watson12
1 NOAA Fisheries, SWFSC, Fisheries Ecology Division, 110 Shaffer
Rd., Santa Cruz, CA 95060 2 435 El Sueno Road, Santa Barbara, CA
93110
3 Environmental Studies Program, University of California Santa
Barbara, Santa Barbara, CA 93106 4 Department of Geography,
University of California Santa Barbara, Santa Barbara, CA 93106
5 ENTRIX Incorporated, 8010 W Sahara Ave, Las Vegas, NV
89117-7927 6 U.S. Geological Survey, Alaska Science Center, 1011
East Tudor Road, Anchorage, AK 99503
7 Native and Anadromous Fish and Watershed Branch, California
Department of Fish & Game, 830 S Street, Sacramento, CA
95814
8 San Diego State University, 5500 Campanile Drive, San Diego,
CA 92182 9 Biological Sciences, San Jose State University, One
Washington Square, San Jose, CA 95192
10 ENTRIX Incorporated, 2140 Eastman, Suite 200, Ventura, CA
93003 11 University of California Davis, 1 Shields Avenue, Davis,
CA 95616 12 The Watershed Institute, California State University
Monterey Bay,
100 Campus Center, Seaside, CA 93955
NOAA-TM-NMFS-SWFSC-394
U.S. DEPARTMENT OF COMMERCE Carlos M. Gutierrez, Secretary
National Oceanic and Atmospheric Administration Vice Admiral Conrad
C. Lautenbacher, Jr., Under Secretary for Oceans and Atmosphere
National Marine Fisheries Service William T. Hogarth, Assistant
Administrator for Fisheries
-
Contents Part 1.
Introduction.....................................................................................................................................................1
1.1. First Goal: Normal Condition as a Reference Point
...................................................................................2
1.2. Second Goal: Identify Populations for Recovery
Planning......................................................................3
1.3. Life-History Plasticity
.....................................................................................................................................4
1.4. Available Information
.....................................................................................................................................6
1.4.1. Data on Distribution and
Abundance....................................................................................................6
1.4.2. Genetic Data
..............................................................................................................................................7
1.4.3. Landscape
Data.........................................................................................................................................9
1.4.4. Climate Data
..............................................................................................................................................9
1.4.5. Data on Stream Discharge
.......................................................................................................................9
Part 2. Identifying the Original Steelhead Populations
.......................................................................................11
2.1. SCCC
Section..................................................................................................................................................11
2.2. NOLA Section
................................................................................................................................................13
2.3. SOLA
Section..................................................................................................................................................14
2.4. A Working Definition of
Population..........................................................................................................17
2.5. Historic Steelhead
Populations....................................................................................................................18
2.6. Three Discrete Populations in the Salinas
System.....................................................................................23
Part 3. Extant Populations
.......................................................................................................................................24
3.1. Artificially Isolated Populations
..................................................................................................................25
3.1.1. Recent Evidence about Relationships of the Isolated
Populations ..................................................25
3.1.2. Recent Evidence about Potential for Anadromy
...............................................................................26
Part 4. Distribution of Potential Steelhead Habitat
..............................................................................................28
4.1. Intrinsic
Potential...........................................................................................................................................28
4.2. Reparameterizing IP using Local Data
.......................................................................................................29
4.3. Preparation of Models for Potential
Habitat.............................................................................................30
4.4. Potential Habitat in the SCCC Section
.......................................................................................................32
4.5. Potential Habitat in the NOLA
Section......................................................................................................36
4.6. Potential Habitat in the SOLA
Section.......................................................................................................40
4.7. Discussion: Key Assumptions and Issues of Interpretation
....................................................................45
Part 5. Assessing Potential Viability of Unimpaired
Populations......................................................................48
5.1. Key Concepts in Viability
Theory................................................................................................................48
5.2. Expectations for the Study Area
..................................................................................................................50
5.3. A Qualitative Ranking
System.....................................................................................................................50
5.3.1. The Population-Size
Assumption.........................................................................................................51
5.3.2. The Habitat-Quantity
Assumption.......................................................................................................56
5.3.3. The Disturbance-Regime
Assumption.................................................................................................58
5.3.4. The Estimation
Assumption..................................................................................................................63
5.4. Potential Viability in the South-Central California Coast
Steelhead ESU.............................................65 5.5.
Potential Viability in the Southern California Coast Steelhead ESU
....................................................67 5.6. Using
the Ranks for Recovery Planning
.....................................................................................................69
Part 6. Assessing Potential Independence of Unimpaired
Populations...........................................................70
6.1. Assumptions and
Analysis...........................................................................................................................70
6.2. An Index of Dispersal Pressure
...................................................................................................................71
6.3. Dispersal Pressure in the South-Central California Coast
Steelhead ESU...........................................74 6.4.
Dispersal Pressure in the Southern California Coast Steelhead ESU
....................................................74 6.5. Using
the Ranks for Recovery Planning
.....................................................................................................76
-
6.5.1. Low-ranked for Both Independence and Potential
Viability............................................................76
6.5.2. Low-ranked for Independence, High-ranked for Potential
Viability ..............................................77
Part 7. Summary
.......................................................................................................................................................78
7.1.
Introduction....................................................................................................................................................78
7.2. Methods
..........................................................................................................................................................78
7.3.
Findings...........................................................................................................................................................79
7.3.1. Original and Extant Populations
..........................................................................................................79
7.3.2. Potential Viability
...................................................................................................................................81
7.3.3. Potential
Independence..........................................................................................................................82
Part 8. Literature
Cited.............................................................................................................................................84
Part 9. Glossary
.........................................................................................................................................................91
Part 10.
Appendices..................................................................................................................................................92
10.1. Evidence for Two or More Populations in the Salinas
Basin.................................................................92
10.1.2. Original Occurrence by
Sub-basin......................................................................................................93
10.1.3. The “Juvenile-Corridor” Hypothesis
.................................................................................................93
10.1.4. The “Hydrologic Forcing” Hypothesis
..............................................................................................93
10.1.5. Other Considerations
...........................................................................................................................97
10.1.6. Three Populations
Perhaps..................................................................................................................97
10.2. The Historic Condition of Mainstem Habitat
..........................................................................................98
10.2.1. The Monte
..............................................................................................................................................98
10.2.2.
Down-cutting.........................................................................................................................................99
10.2.3. Formerly Perennial Flow
.....................................................................................................................99
10.2.4. Probable Historic Baseline
Conditions.............................................................................................100
10.3. Historic Accounts of O. mykiss in the SOLA Section of the
Study Area.............................................101 10.3.1.
Santa Ana
River...................................................................................................................................101
10.3.2. San Gabriel River
................................................................................................................................102
10.3.3. Los Angeles River
...............................................................................................................................102
10.3.4. Context
.................................................................................................................................................102
10.4. A Note on Sources of
Information...........................................................................................................104
10.5. Gauge Data for Inland/Coastal
Comparison..........................................................................................105
10.6. Limiting Habitats and Population
Size...................................................................................................107
10.6.1. Ocean
Conditions................................................................................................................................107
10.6.2. Spawning migrations.
........................................................................................................................107
10.6.3. Spawning habitat.
...............................................................................................................................107
Part 11. Color Plates
...............................................................................................................................................109
Typographical errata on title page and pp. 84-90 corrected in
this version (1 Nov. 2006)
-
Acknowledgments
We wish to thank M. Capelli and M. Larson for their support of
this work; S. Cooper and D. Jacobs for com-menting on an earlier
draft; and our various research and management institutions for
their administrative support. Finally, we wish to particularly
thank the myr-iad scientists, citizens, and steelhead enthusiasts
whose observations formed the basis for this retrospective look at
steelhead population structure along the south-central and southern
California coast.
-
1
Part 1. Introduction
The aim of the Federal Endangered Species Act (ESA) is to
recover species that would other-wise go extinct, and to this end
it requires the Fed-eral government to prepare recovery plans. A
re-covery plan outlines a strategy for lowering ex-tinction risk to
an acceptable level, and has two components: a technical part and a
policy part.
The technical part evaluates information on the species itself,
and especially the changes in abundance, distribution, habitat
condition, etc., that would reduce the extinction risk. The policy
part determines which of the risk-reducing changes are desirable
and feasible, outlines the steps necessary to bring them about, and
estimates their cost. For West Coast salmon and steelhead, these
two parts are formally labeled “Phase I” technical recovery and
“Phase II” implementation by the National Marine Fisheries
Service.
This report concerns Phase I, and applies a formal evaluation
framework developed else-where (McElhaney et al. 2000, Bjorkstedt
et al. 2005) to the problem of delineating Oncorhynchus mykiss
populations in the South-Central/Southern California Coast recovery
domain. These popula-tions inhabit the set of coastal river basins
encom-passed by the Pajaro basin in the north and the Tijuana basin
in the south, hereafter referred to as the study area (Figure 1).
According to a coast-wide status review of steelhead described by
Busby et al. (1996), the study area is inhabited by two
Evolutionarily Significant Units (ESUs) of O. mykiss.
The ESU concept comes from Waples (1991), who considered a group
of O. mykiss to comprise an ESU if 1) they were substantially
reproduc-tively isolated from other conspecific population units,
and 2) they represented an important com-ponent of the evolutionary
legacy of the species. The distinct Mediterranean ecology of the
study area, and its division into two faunal provinces on either
side of Point Conception, led Busby et al. (1996) to designate two
important components of the evolutionary legacy of the species,
with geo-graphic ranges as in Figure 1; some genetic
con-siderations also played a role in this analysis.
These units were named the South-Central Cali-fornia Coast
Steelhead ESU and the Southern Cali-fornia Coast Steelhead ESU, and
we follow this convention here.
The steelhead (anadromous) portion of each ESU is currently
listed on the US Endangered Spe-cies List as a threatened or
endangered Distinct Population Segment, or DPS (Federal Register
70: 67130 [2005] & 71: 834 [2006]). Anadromous fish are those
that spend some part of their adult life in the ocean, in contrast
to non-anadromous fish that spend their entire lifecycle in
freshwater systems. Both forms occur in the study area. Formally,
the steelhead DPS of O. mykiss includes only those individuals
whose freshwater habitat occurs be-low impassible barriers, whether
artificial or natu-ral, and which exhibit an anadromous
life-history. Operationally, distinguishing listed from non-listed
O. mykiss can rely on features such as rela-tive size, smolting
behavior, feeding activity, length of migratory movement, and
number of eggs produced by females, but cannot rely on
re-productive isolation between the anadromous and non-anadromous
forms (See Federal Register 71: 834 [2006]). Thus, listed
anadromous forms and unlisted non-anadromous forms can co-exist in
the same ESU or even the same population, and in-deed there is
evidence suggesting both kinds of co-existence (or polymorphism)
are present in the
Figure 1. The study area and the geographic ranges of its
steelhead ESUs.
-
2
study area. This means that discussions of popula-tion
delineation and distribution necessarily in-volve considering both
forms of O. mykiss jointly in natural units of ecological
organization (popula-tions and ESUs).
The authors of this report are members of a Technical Recovery
Team (TRT), convened to ad-vise NMFS on technical aspects of
recovery in the study area. This report has two goals: to describe
the normal (reference) condition of each ESU; and to identify
existing and potential populations of steelhead that could form the
basis for recovery.
It should be noted at the outset, however, that these two goals
are burdened with numerous un-certainties and judgment calls on the
part of the authors. The uncertainty stems from several
inter-acting factors:
1) The extremely large and heterogeneous planning area,
comprising the south-west range limit for the species.
Environmental heterogeneity appears to constrain the distribution
of the species at a number of spatial scales, making the task of
describing this distribution somewhat complex.
2) Most of the information about the species in the study area
comes from anecdotal reports (descriptive in nature) or from
studies conducted at restricted spatial scales (individual reaches,
or at best, large sections of individual watersheds).
3) The task of delineating populations and characterizing
recovery potential is largely reliant on quantitative data samples
from across the planning domain. Since such information is
un-available, we are confined to the less satisfactory exercise of
A) applying simplistic yet uniform methods over large spatial
extents, and B) describ-ing existing small-extent studies, and
making un-certain inferences of their implications for the lar-ger
ESU. For the most part, these two approaches lack the level of
quantitative description that is necessary for making concrete
recommendations.
There is a natural tension between the simple broad-extent,
coarse-resolution mode of analysis and the small-extent,
high-resolution mode of analysis alluded to above. In describing
both modes, we hope to provide a useful reference for recovery
efforts, and to clarify the relative utility of future research on
O. mykiss that might be con-ducted in the study area.
1.1. First Goal: Normal Condition as a Reference Point
Recovery is defined by the National Marine Fisheries Service as
“the process by which listed species and their ecosystems are
restored and their future is safeguarded to the point that
protections under the ESA are no longer needed” (NMFS 2004). Such
restoration first requires a description of the normal condition to
which the species is to be restored. For ESU structure, normal
condition is most conveniently described in terms of individ-ual
populations: where they are located, how resil-ient each one is to
extinction, and so forth. In this context, “normal condition” can
have at least two meanings, the simplest being the original
popula-tion structure of the ESU prior to the arrival of non-native
Americans. This concept has three problems in our case: 1)
settlement-era accounts of steelhead are extremely sparse (Titus et
al. 2003); 2) the abundance of steelhead during the settlement era
may have been unusually high due to the pre-ceding demise of Native
Americans (a key preda-tor) from small pox (see Keeley 2002b), and
3) the climate of southern California has been changing, getting
wetter and warmer since the ending of the “Little Ice Age” in the
19th Century (Millar and Woolfenden 1999; Haston and Michaelsen
1997; Scuderi 1993). O. mykiss are probably especially vulnerable
to climate change in the study area, as it contains their southern
range limit. Presumably the species is near the limits of its
tolerance for warm or dry conditions, and small changes in cli-mate
may well translate to large changes in poten-tial steelhead
distribution. This would cause the 19th Century to be a misleading
reference point.
The other meaning of “normal condition“ would be the
hypothetical present-day state of each ESU if non-native Americans
had had no sig-nificant impact on the fish. Though a hypothetical
construct, this concept of “unimpaired population structure” is in
many ways more useful for recov-ery because it can be studied using
data collected from relatively unimpaired stream systems in the
present climate1. Moreover it is directly relevant to
1 Unimpaired is a relative term, since the natural function of
most and perhaps all streams in the study area has been af-fected
to some degree by human immigrants
-
3
recovery under current climatic patterns. How-ever, “unimpaired”
should not be taken to mean “static,” as unimpaired condition
involves dy-namical regimes that are characteristic of a given
ecosystem (such as terrestrial fire regimes or ma-rine ecosystem
responses to decadal climate pat-terns (e.g. Mantua and Hare
2002).
Here, the overall focus will be on the unim-paired population
structure rather than the origi-nal structure, since it is the most
relevant reference point for recovery planning. However, due to
on-going climate change, the difference between un-impaired
structure today and 50 yrs hence could be quite large—perhaps much
greater than the difference between now and 200 yrs in the
past.
In addition to unimpaired structure and origi-nal structure,
there are two other useful reference points, current structure of
the ESU, and ESU vi-ability (Table 1). The level of recovery
necessary to achieve ESU viability is not addressed in this
re-port, but will be discussed at length elsewhere.
1.2. Second Goal: Identify Populations for Recovery Planning
In a scientific review focused on recovery planning for west
coast salmonids, McElhany et al. (2000) concluded that independent
viable popula-tions are the basic components of a viable ESU.
Independence and viability are defined thus:
“A viable salmonid population is an independent popu-lation of
any Pacific salmonid that has a negligible risk of extinction due
to threats from demographic variation (random or directional),
local environmental variation, and genetic diversity changes
(random or directional) over a 100-year time frame” (McElhany et
al. 2000:2) “The crux of the population definition used here is
what is meant by ‘independent.’ An independent population is any
collection of one or more local breeding units whose population
dynamics or extinction risk over a 100-year time period is not
substantially altered by exchanges of individuals with other
populations.” (McElhany et al. 2000:3)
To use these concepts, it is necessary to divide each ESU into
individual populations and assess the independence and viability of
each.
The reason for taking these steps is that certain populations
may not be viable even in their origi-nal or unimpaired state. This
could occur for ex-ample in small coastal basins that do not
provide enough habitat to support a large, persistent steel-head
run, and rely on periodic immigration for long-term presence. A
possible example is the steelhead in Topanga Creek, which though
com-mon in the 1960s were not observed during the 1980s and most of
the 1990s, but later reappeared near the end of the century (Dagit
and Webb 2002). A reasonable interpretation of these data is that
the local population went extinct, but later was re-established by
steelhead from elsewhere. Topanga Creek is a small stream system,
pre-sumably with a small carrying capacity for steel-head. In
general, small populations are expected to turnover in the manner
of Topanga Creek steel-head, so the data are not surprising.
However, one would not want to base a recovery plan on popu-lations
that are, even if completely recovered, so small or so unstable
that they are vulnerable to local extinction. Thus, the second goal
of this re-
Table 1. Reference points in ESU recovery.
Term Definition
Original population structure
The population structure of the ESU at the arrival of permanent
settlers of European descent (c. 1769 – 1850)
Unimpaired population structure
The hypothetical present-day structure of the ESU if non-native
Americans had had no significant impact on the fish.
ESU viability
The hypothetical state(s) in which extinction risk of the ESU is
neg-ligible.
Current population structure
The current population structure of the ESU (c. 1970 –
2005).
-
4
port is to identify populations that have the inher-ent
potential to be well-buffered from extinction if restored to their
unimpaired state. Consequently, our primary tasks in this report
are: 1) Identify all the original steelhead populations
in the study area, and determine which ones are extant;
2) Delineate the potential unimpaired geographic extent of each
population;
3) Estimate the potential viability of each popula-tion in its
(hypothetical) unimpaired state; and
4) Assess the potential demographic independ-ence of each
population in its unimpaired state.
1.3. Life-History Plasticity Before going further, it may be
informative to
review what is known about life-history plasticity of the
steelhead in California, as it is somewhat complex and intricate
but very key to understand-ing the rest of the document.
These fish are flexible in their approach to life—they can
complete their life cycle completely in freshwater, or they can
migrate to the ocean after 1 – 3 yrs, and spend 2 – 3 years in the
marine environment before returning to spawn. The fish pursuing the
former life history trajectory is com-monly called a rainbow trout
in our study area, and the latter is called a steelhead, but it has
be-come clear in recent years that this terminology is misleading
in its simplicity.
For one thing, rainbow trout sometimes have steelhead as
progeny, and vice versa. These facts have been demonstrated by
studying the otolith microchemistry of O. mykiss. Otoliths are
small ear bones that lay down growth increments that are the
ichthyological equivalents to tree rings. More-over, the isotopic
composition of the increments depends on whether the fish inhabited
fresh or salt water at the time the increment was laid down. As a
result, mass-spectrometry can be used to reconstruct the isotopic
timeline, and therefore the freshwater-marine timeline, of a given
fish’s life history. The isotopic composition of the otolith
primordium is determined by the habitat of the mother, and this
allows a comparison of parent-
offspring life histories. Zimmerman and Reeves (2000) used
techniques such as this to uncover oc-casional life-history
“switching” in certain O. mykiss populations in Oregon. The
steelhead in our study area have not yet been examined in this way,
but numerous anecdotes indicate that life-history switching is
probably widespread. We do not know what cues it.
For another thing, there is a third group of life history
strategies, that we here call “lagoon-anadromous.” Bond (2006),
working at a study site in northern Santa Cruz County, has recently
shown that each summer a fraction of juvenile steelhead
over-summered in the estuary of their natal creek. Like elsewhere
in California, this estu-ary was cut off from the ocean during the
summer by the formation of a sandbar spit, and thus is more
properly referred to as a seasonal lagoon. Bond (2006) showed
unequivocally that juvenile steelhead do very well if they
over-summer in the lagoon—many grow fast enough to migrate to the
ocean their first year, and most enter the ocean at a larger size
than fish coming from the freshwater portion of the stream system.
Large size enhances survival in the ocean, and thus the
lagoon-reared fish tend to be disproportionately represented in the
adult spawning population (Bond 2006).
Within each of the three basic life-history groups (freshwater
resident, lagoon-anadromous and fluvial-anadromous), there is
additional varia-tion: Juveniles may spend 1 – 3 yrs in freshwater,
1 – 2 yrs in the lagoons, and adult steelhead may spend from 2 – 3
yrs in the ocean before returning to spawn. Finally, unlike other
Pacific salmon, some adults survive their first spawning and
re-turn to the ocean to wait for next year. A graphic overview of
this life-history diversity, along with some of the specialized
terminology, is given in Figure 2.
On top of all this, there are examples of finer-scale habitat
switching, such as multiple move-ments between lagoons and
freshwater in the course of a single summer; and also so-called
“ad-fluvial” populations that inhabit reservoirs but spawn in
tributary creeks. O. mykiss are flexible in their approach to
life.
-
5
Figure 2. A synopsis of life-history trajectories believed to
occur in the study area. Relative frequency of each trajectory is
not known.
-
6
1.4. Available Information We briefly review the available data
that bear
on the distribution, abundance, and potential habitat of
steelhead in the study area, because these data are the ultimate
basis for identifying and characterizing populations.
Usefulness of the available data are based on relevance,
credibility, and geographic consistency. A discussion of relevance
we defer to the next sec-tion. Credibility of information is
considered rela-tively high when information has been published and
peer-reviewed. Much other information is available as reports that
have not necessarily been through a formal peer-review process or
even made publicly available. Due to paucity of infor-mation we
found it useful to cite many such re-ports, but were faced with the
task of judging their credibility. To do so, we adapted
recommenda-tions from Walton (1997), a practical philosophical
treatise on judging the validity of expert opinions (for more
detail, see §10.4, p. 104).
The geographic consistency of a given source of information is
what allows the broad-extent, coarse-resolution analyses that we
alluded to in the introduction.
1.4.1. Data on Distribution and Abundance Data on run size—the
number of adult steel-
head spawning in a particular stream during a particular
winter—would be extremely relevant but are almost non-existent for
the study area (as well as most other parts of California). The
notable exception is the Carmel River, for which run size has been
monitored since 1964 at the fish ladder on San Clemente Dam (with a
gap from 1978 – 1987; see Snider 1983, Williams 1983, and Mon-terey
Peninsula Water Management District web-site2). The count is
incomplete because some pro-portion spawns downstream of the dam;
expert opinion puts the proportion somewhere in the range of 10% -
50%.
There are also accounts of “typical” historical run size for
many of the domain’s largest basins. The accounts are generally
based on expert opin-ion rather than data (Boughton 2005), and
there is
2 http://www.mpwmd.dst.ca.us
little agreement among today’s experts as to their accuracy. For
the smaller basins, and the basins south of Los Angeles, there are
usually no credible estimates of historical run size at all.
A number of single-basin studies of fish dis-tribution have been
made in recent years, includ-ing Smith’s (2002) summary of the
upper Pajaro River system; Alley’s (2001) monitoring on Santa Rosa
Creek in San Luis Obispo County; Payne and Associates’ (2001, 2004)
survey around Morro Bay and San Luis Obispo Creek; the informative
sur-vey of the Salinas basin by Casagrande et al. (2003); the
thesis of Douglas (1995) on O. mykiss in the Santa Ynez basin;
Kelley’s (2004) assessment of the Santa Clara basin; the assessment
by Stoecker and the CCP (2002) on the Santa Barbara Coast and
Stoecker and Stoecker (2003) on the Sis-quoc River; Allen’s (2004)
assessment of the Ven-tura River basin; a study of steelhead
habitat in the Santa Monica Mountains (California Trout 2005);
Kelley and Stoecker’s (2005) assessment of recov-ery opportunities
in the Santa Clara River, and Spina and Johnson’s (1999)
examination of Solstice Creek in the Santa Monica Mountains (for
more information, refer to the descriptive summary starting on page
11). Though these types of studies provide insight into the status
of particular basins, they do not cover all steelhead-bearing
watersheds and are thus not definitive. Also, since each has a
unique set of goals and study design, they are not always
comparable.
A simple but useful type of information is oc-currence data,
also called “presence-absence” data. Occurrence data are sparse for
pre-history; the authoritative reference is Gobalet et al. (2004),
who used archaeological records to establish the occur-rence of
Oncorhynchus mykiss in 25 coastal locali-ties between San Francisco
and Mexico. The southern-most was Los Peñasquitos Creek in San
Diego County, confirming that the species oc-curred at least this
far south prior to European settlement.
Titus et al. (2003) have made a concerted effort to track down
occurrence data in the historical record. Most of this record
consists of field notes from CDFG biologists active in the early
20th Cen-tury. Also, Sleeper (2002) and Franklin (1999) gathered
oral accounts from elderly citizens of
-
7
Orange/San Diego Counties and the upper Salinas Valley,
respectively. These three manuscripts summarize eyewitness accounts
of steelhead, mostly from the early-to-mid 20th Century, by which
time much environmental change had oc-curred resulting from the
decline of the large Na-tive American population and the arrival of
Euro-peans in the 18th and 19th centuries (See §10.2 on p. 98).
Even so, the accounts provide much credible information about the
geographic distribution of steelhead at the resolution of named
creeks. Frank-lin (1999) and Sleeper (2002) had an emphasis on
observations of adults; Titus et al. (2003) tended to emphasize
juveniles. Recently, Kelley and Stoecker (2005) provided a table
summary of early steelhead reports within the Santa Clara River
ba-sin.
For the immediate past, there are numerous sources of occurrence
data. The basin-scale studies mentioned above have numerous data,
assignable to particular reaches on particular dates. Boughton et
al. (2005) made an assessment of occurrence across the entire
domain; reports such as those by Payne and Associates (2001, 2004)
contain useful accounts; and occurrence data are also preserved in
the research collections of the California Acad-emy of Science and
the Los Angeles County Mu-seum. Many of the occurrences specify
latitude and longitude; or give detailed locality descrip-tions.
These data, though collected rather haphaz-ardly through the years,
appear to be the best fish data we have in terms of overall
credibility, geo-graphic extent, and geographic resolution.
1.4.2. Genetic Data Since the late 1980ʹs, a number of studies
have
been conducted to elucidate the genetic structure of steelhead
populations in the study area. Early studies used
electrophoretically detectable protein differences (allozymes).
More recently, studies have employed molecular genetic analyses,
assay-ing variation in mitochondrial DNA (mtDNA) sequence, and
variation in tandem-repeat copy number of microsatellite loci.
Berg and Gall (1988) surveyed 24 polymorphic allozyme loci from
populations throughout Cali-fornia, including a small number of
populations
from the study area. They discovered considerable variability
among California populations, but did not discern a clear
geographic pattern to the varia-tion.
Busby et al. (1996) report a large-scale study of 51 allozyme
loci in 113 populations, including 22 from California, four of
which were specifically from the study area. A high level of
genetic vari-ability was found in the California coastal
popula-tions. The most remarkable feature of the data was a cline
in frequency of the “70” allele at the fruc-tose–biphosphate
aldolase-3 (FBALD-3) locus. The allele occurred either rarely, or
not at all, in steel-head samples from coastal Oregon and the
Klamath Mountains province, but its frequency in the samples
increased north to south down the California coast, and was the
only allele present in the southernmost sample at Gaviota Creek.
Busby et al.(1996) noted that finding an allozyme allele fixed in
some populations, but entirely absent in others, is unprecedented
in salmon, except when comparing populations at the extreme ends of
their ranges.
Over all loci, however, there was not a clear pattern of
population affiliation among the popu-lations south of the Eel
River. For example, a mul-tidimensional scaling plot showed that
the two southernmost populations in the study (Arroyo Hondo and
Gaviota Creek) were not closely re-lated to each other even though
they are located near one another and are divergent from most other
California populations. This was attributed to four possibilities:
1) the extreme and variable habitat conditions of southern
California promote local adaptation, and hence isolation, between
southern steelhead populations, 2) increased reli-ance on
freshwater residency and maturation in the south leads to increased
isolation between populations, 3) small population size allows
ge-netic drift–the change of allele frequencies due to the random
nature of genetic inheritance–to pro-ceed more rapidly in the
southern populations, and 4) haphazard sampling (i.e. non-random,
non-systematic sampling of fish in space and time).
In the 1990s, Nielsen began a series of investi-gations into the
molecular genetic diversity and biogeography of steelhead in our
study area. Niel-sen et al. (1994) assayed genetic variation in
-
8
mtDNA and a single microsatellite locus in 468 coastal O. mykiss
sampled from 31 populations throughout California. Allele
frequencies differed enough between populations to reject the
hy-pothesis that steelhead throughout southern Cali-fornia are
freely interbreeding. Nielsen et al. (1994) offered two
explanations for this: 1) genetic drift has caused populations in
southern California to differ from one another and from the rest of
the California populations, or 2) the southern steel-head are
descended from an ancient lineage that survived the Pleistocene in
a refugium in the Gulf of California. The authors noted that the
data were insufficient to reject either explanation, but pre-dicted
that if explanation 2 were true, then a high degree of genetic
diversity should be observed in our study area.
Nielsen et al. (1997) compared genetic diver-sity in mtDNA and
three microsatellite loci in O. mykiss from five habitats with
varying degrees of hatchery influence and accessibility to the
ocean. Samples were drawn from streams with and with-out access to
the ocean, reservoirs, and hatcheries, and from sea-run adults and
outmigrating smolts (the anadromous group). Based on the presence
of rare haplotypes, mtDNA diversity was found to be highest among
the anadromous fish (however, this result may be an artifact of the
small number of anadromous fish sampled), and lowest among the
hatchery trout. A similar pattern was observed at the three
microsatellite loci.
Additionally, certain “uniquely southern” haplotypes absent in
rainbow trout hatchery strains occurred at moderate frequency in
rainbow trout from freshwater habitats–both with and without ocean
access–throughout the study. This suggested that some rainbow trout
populations in southern California, despite years of stocking with
hatchery strains, still possess genetic heritage from wild southern
steelhead. It was pointed out, how-ever, that rainbow trout from
streams with open access to the ocean were more closely related to
the anadromous fish than were fish from closed habitats or
reservoirs, suggesting that trout that still have access to the
ocean may retain a greater degree of southern steelhead heritage.
While these are interesting suggestions, the authors empha-sized
that studying direct introgression between
hatchery fish and remnant southern steelhead populations is
difficult because of their shared evolutionary history, and, hence,
genetic similarity between coastal steelhead and some hatchery
populations.
Nielsen et al. (1998) documented D-loop varia-tion in mtDNA from
5 species of Oncorhynchus and reported that coastal O. mykiss
carried the highest haplotype diversity (number of haplo-types)
found in the study. However, this could be an artifact of sample
size: their sample sizes were largest for coastal steelhead and no
attempt was made to account for sample size in the number of
haplotypes observed. They also found that south-ern steelhead and a
trout from Mexico comprised the most genetically separated O.
mykiss popula-tions. However, this claim was based on a single
genetic locus (the mtDNA) and had low statistical support. As the
authors noted, “population differ-entiation based on putatively
neutral genetic variation holds only speculative value in drawing
evolutionary inference at the fine scale of intras-pecific or
subspecific analyses.”
In Nielsen (1999), 11 microsatellites were typed from a small
number of anadromous fish collected over 8 years from southeast
Alaska to Malibu Creek. Several alleles were recorded in northern
and southern California populations that were not previously
reported in populations of steelhead in Washington. In fact, at
nine of the 11 loci, alleles were observed in California that were
outside the size range of alleles observed in Puget Sound.
Nielsen (1999) deemed it unlikely that wider allele size ranges
would occur in California if steelhead survived the late
Pleistocene in a single northern refugium, and then colonized
rivers to the south in California, Thus, she argued that “we are
left with one alternative to explain the unique genetic diversity
observed...the vicariance model of genetic variation,” and that
“Perhaps some of the genetic diversity in southern steelhead
repre-sents lineage effects from populations that evolved from a
Gulf of California refugium, rather than reflecting particular
processes in a marginal popu-lation with common ancestry from a
Beringia refugium” (p. 456). This is a compelling hypothe-sis, and
accords somewhat with Behnkeʹs (1992)
-
9
view that “coastal rainbow trout diverged from the redband
line...possibly during the late Pleisto-cene...and perhaps in
California” (p. 20), but it re-mains unknown whether the data
presented in Nielsen (1999) are sufficient to reject the
possibility that southern steelhead may, in fact, have de-scended
with northern populations from a com-mon refugium. There is not a
standard statistical test for such a proposition.
Several other analyses in Nielsen (1999) gave a mixed picture of
the distinctness of southern steelhead: first, a neighbor-joining
tree based on delta-mu, microsatellite distances supported
sepa-ration of southern steelhead from the rest of the populations
in the study with the low (not statisti-cally significant)
bootstrap value of 57%. The weak statistical support may be due to
the small sample sizes. And second, evidence was presented of
iso-lation by distance, but the signal was diminished with the
southern steelhead included in the data, possibly because the
southern steelhead are not in genetic equilibrium due to a recent
genetic bottle-neck.
1.4.3. Landscape Data Environmental science now has available to
it a vast array of mapped information, deployed as computerized
Geographic Information Systems (GIS). This kind of data is usually
produced by some combination of remote sensing, field studies, and
geographic modeling. A classic example of a such information is the
Digital Elevation Model (DEM), which represents the Earth’s surface
as a grid of 30m x 30m cells (sometimes 10m x 10m), and specifies
the mean elevation of each cell. Older DEMs are basically digitized
versions of USGS topographic maps; new DEMs are gener-ated from
NASA’s Shuttle Topography Radar Mis-sion (STRM). A DEM is the basis
for many derivative data-sets that are relevant to steelhead
recovery. For example, an algorithm can use the topographic
information to identify stream networks and automatically map them.
This is useful for high-order streams that are not well portrayed
in the original USGS maps, although the algorithm does not perform
particularly well in flat areas, such as
alluvial valleys. Two other applications useful for steelhead
are the use of DEMs to delineate catch-ment basins; and to estimate
valley width. The latter is the lateral area around a stream
channel in which the channel can migrate over time due to erosion
and depositional processes. There are numerous other sorts of
geoenvi-ronmental data available, describing land cover, geology,
etc. Those geoenvironmental datasets that are relevant to steelhead
ecology are useful, be-cause they are generally credible and have
broad geographic extent. However, in many cases the resolution of
the data can be limiting.
1.4.4. Climate Data Daly et al. (1994) describe a model for
map-
ping climate data in a GIS framework. In particu-lar, based on a
mechanistic understanding of how broad-scale climate patterns
interact with topog-raphy, they developed a model that allowed them
to use data from the US network of weather sta-tions
(precipitation, temperature) to create com-plete maps of certain
climate norms, such as mean annual temperature, mean annual
precipitation, mean monthly precipitation, and so forth. Since we
expect that both temperature and precipitation are key limiting
factors for steelhead distribution in our study area, we expect
these datasets to be useful. They are available online at the
Spatial Climate Analysis Service3. Some pertinent exam-ples of the
data are in Plate I through Plate VI at the end of this report.
Detailed overviews of this approach to climate modeling can be
found in Daly et al. (1994, 2001 and 2002).
1.4.5. Data on Stream Discharge The United States Geological
Service main-
tains data from numerous stream gauges within the study area,
some of which provide useful his-torical context4. For example, one
gauge has been in continuous operation on the Arroyo Seco, a
tributary of the Salinas River, since 1901. These datasets consist
of mean daily discharge for the period of record at each gauge.
3 http://www.ocs.orst.edu/prism/ 4
http://waterdata.usgs.gov/ca/nwis/nwis
-
10
Needless to say, water flow is fundamental to the occurrence of
O. mykiss, and in southern Cali-fornia it is so variable that it
cannot be taken for granted. Gauges are widespread in the study
area, but their distribution is irregular and hence their
geographic consistency is only moderate. Even so, these data are
potentially useful in three respects. The first is that discharge
data describe, to some degree, what is occurring upstream and
down-stream. In this sense gauge data have greater geo-graphic
extent than just the point at which the data were collected.
The second respect in which the data are use-ful is for fitting
models that are geographically consistent. For example, Burnett et
al. (2002) used gauge data in Oregon to predict discharge in
un-gauged reaches, so as to assess the potential dis-tribution of
coho salmon (Oncorhynchus kisutch). The predictors in their model
were geoenviron-mental coverages with high geographic
consis-tency—namely, contributing watershed of each reach in a GIS
stream network, and a coverage of mean annual precipitation. To the
extent that dis-charge can be regressed on these two datasets, they
can be combined with USGS discharge data to make a predictive map.
The map is spatially consistent, relevant, and credible because its
accu-racy can be assessed using standard statistical procedures.
Beighley et al. (2005) describe a more refined approach for
constructing rainfall-run-off models for a portion of the study
area. The third respect in which the data are useful is that they
provide information on migratory ac-cessibility for the fish. In
arid regions where rain-fall is variable within and among years, it
is thought that discharge is so variable that it does not provide
reliable access for steelhead migrating to or from the ocean. USGS
gauge data provide a means to compare the migration reliability of
streams empirically. When using gauge data to interpret patterns of
discharge, it is important to recognize some limitations of stream
gauges. One key limitation is
that stream gauges omit data on groundwater or hyporheic flow.
In some cases losses to, or gains from, groundwater can be quite
substantial, espe-cially at low flow (Figure 3). In addition, many
USGS gauges are not designed to accurately re-cord low flows (2 – 5
cfs), and because of irregular maintenance some are not always
operable. In consequence, estimates of low flows are some-times
biased low or erroneously reported as zero.
Figure 3. Daily mean discharge during the steelhead migration
season (Jan – May), for two gauges on the Arroyo Seco, a steelhead
stream in the Salinas basin. The old gauge (11152000) is situated
where the Arroyo Seco leaves the Sierra de Salinas and enters the
broad alluvial Salinas Valley. The other gauge (11152050) is
located 17 km downstream, near the confluence with the mainstem
Salinas. Between the two sites, a signifi-cant proportion of
discharge is lost under low-flow conditions, as indicated by the
downward-curve of the cloud of datapoints. Presumably the “lost”
water en-ters the Salinas aquifer underlying the alluvial valley.
Data are in cubic feet per second.
-
11
Part 2. Identifying the Original Steelhead Populations
The stream systems of the study area are ar-rayed as a string of
173 coastal basins, bounded by the Pajaro Basin in the north and
the Tijuana Basin in the south. This study area encompasses about
half the California coast, a span that includes red-wood forests,
oak savannah, chaparral and high desert. The study area is rather
large and has ex-ceptional ecological diversity. For ease of
presen-tation, we divide it into three sections of about equal
geographic area (Figure 4). The SCCC sec-tion is inhabited by the
South-Central California Coast Steelhead ESU. The NOLA and SOLA
sec-tions are both inhabited by the Southern California Coast
Steelhead ESU, although as we shall see there is more steelhead
information about the NOLA section. Below we give an overview of
each section.
2.1. SCCC Section The SCCC section is inhabited by the
South-
Central California Coast Steelhead ESU. The sec-tions’ two
northernmost basins—the Pajaro River system and Salinas River
system—are also the two largest basins in the entire study area
(Figure 5). A distinctive feature of both is their penetration to
the interior of the coastal mountain ranges, which is significantly
more arid and seasonal than the coastal slopes (see Plate I, Plate
IV). Another dis-tinctive feature is the long alluvial lower
stretches of the mainstem Pajaro and Salinas. It is suspected that
during severe droughts, these lower channels may have caused
problems for fish passage (espe-cially for smolts) and therefore
were a source of stochasticity that made inland populations less
stable.
A segment of the Pajaro system drains the southern end of the
Santa Cruz Mountains, an area of dense redwood forest and cool
mountain creeks. For more information on the Pajaro, see Stanley et
al. (1983) and Smith (2002); for the Salinas see Casagrande et al.
(2003).
South from the Salinas estuary, a notable sys-tem is the Carmel
River basin, larger than the coastal systems of the Big Sur to the
south, but
nowhere near the size of the neighboring Salinas Basin. The
Carmel is a well-known steelhead stream (Snider 1983), and
continues to be actively managed by various entities. Unlike the
systems of the Big Sur to the south, the lower reaches of the
Carmel have an alluvial character somewhat like the Pajaro and
Salinas.
The Big Sur coastal area is south of Carmel and has a moderate
climate—cool foggy summers and warm wet winters. Vegetation
consists of oak parklands and chaparral, but also stands of
Doug-las fir and small pockets of redwoods. The stream systems
occur as numerous small coastal basins draining the steep
Pacific-facing slopes of the Santa Lucia Mountains. Along with the
southern Santa Cruz Mountains, the central Big Sur is one of the
two distinctly wet places in the study area (Plate IV).
Figure 4. Short-hand acronyms for sections of the study area
discussed in the text. SCCC = South-Central California Coast; NOLA
= North of Los Angeles; SOLA = South of Los Angeles.
-
12
Figure 5. SCCC Section, showing principal streams, towns, and
mountain ranges.
-
13
In the vicinity of the Monterey-San Luis Obispo county line, the
steep coastal slopes of the Big Sur give way to marine terraces as
the Coast Range heads slightly inland. This pattern of ter-race
backed by mountains is typical of the coastal streams of San Luis
Obispo County. Examples are Santa Rosa Creek in north county, San
Luis Obispo Creek in central county, and Arroyo Grande in south
county (Alley 2001, Payne and Associates 2003). Arroyo Grande was
mentioned by David Starr Jordan as a popular angling spot for O.
mykiss in the early 20th Century (Titus et al. 2003). All these
systems tend to be slightly larger than those of Big Sur and
penetrate further inland; they also differ in that their lower
mainstems tend to be low-gradient channels through raised marine
terraces. To the extent that these lower mainstems retain perennial
flow better than the upper tribu-taries, they serve as important
over-summering habitat for juvenile steelhead (Payne and
Associ-ates 2003). The Salinas basin is large, and its magnitude is
perhaps brought home by the fact that its headwa-ters are adjacent
to the headwaters of the Arroyo Grande. Several significant
tributaries of the Salinas River drain the backside of the Big Sur,
each with a watershed area comparable to that of the Carmel River.
The furthest north is the mis-named Arroyo Seco, which joins the
mainstem Salinas near the ruins of the old Mission Soledad. Further
upstream are the paired Nacimiento and San Antonio Rivers, both of
which join the main-stem at the Camp Roberts Military Reservation.
These systems are true perennial rivers, in contrast to the desert
washes that drain the eastern side of the Salinas Valley.
Casagrande et al. (2003) provide a useful overview of the Salinas
system.
2.2. NOLA Section So far we have been describing basins
inhab-
ited by fish of the South-Central California Coast Steelhead
ESU. Starting with the Santa Maria sys-tem, the basins are
inhabited by the Southern Cali-fornia Coast steelhead ESU. It is
useful to divide this area into a “north-of-Los-Angeles” section
and a “south-of-Los-Angeles” section (NOLA and SOLA,
respectively).
The most northerly basin in the NOLA section is drained by the
broad Santa Maria River, run-ning past the town of the same name
(Figure 7). The Santa Maria River itself is a relatively short
connection between the ocean and its two major tributaries—the
Sisquoc and Cuyama Rivers. Both of these systems, as well as the
large Santa Ynez system just to the south, drain the steep slopes
of the Transverse Ranges before running through wide alluvial
valleys to the ocean. Their headwa-ters drain the coolest, wettest
area in the NOLA section (Plate II, Plate V), the rugged montane
highlands around Monte Arido. Stoecker and Stoecker (2003) give an
overview of steelhead in the Sisquoc system, and Douglas (1995) and
Car-panzano (1996) describes distributional studies of O. mykiss in
the Santa Ynez system.
South of the Santa Ynez River mouth, the coast makes a
right-angle turn to the east at Points Arguello and Conception.
From here to Ventura, the Santa Barbara coast is drained by a set
of small, south-facing coastal basins. These systems all have their
headwaters in the Santa Ynez Moun-tain range that parallels the
coast at this point; their lower sections run through the small
coastal terrace sandwiched between the ocean and the range. One
noteworthy exception to the pattern is Gaviota Creek, which
actually penetrates the Santa Ynez Mountains and drains a small
part of its north slope. Stoecker and the CCP (2002) pro-vide a
useful introduction to steelhead in the Santa Barbara coastal
area.
Continuing down the coast, the pattern of very small basins is
interrupted near the coastal town of Ventura, which is flanked on
either side by the mouths of two large and well-known steel-head
rivers. The first of these is the south-running Ventura River,
whose headwaters drain the south slopes of the same cool and wet
Monte Arido highlands drained on the west by the Santa Ynez and
Sisquoc Rivers (Plate II). The second is the west-running Santa
Clara River, which drains a large and arid area stretching all the
way to Sole-dad Pass just south of Palmdale.
The available evidence suggests that steelhead have been limited
to the western part of the Santa Clara basin (Kelley 2004).
Noteworthy in this part of the basin are two large
tributaries—Sespe and
-
14
Piru Creeks--that arc around to the west, draining the
north-west slopes of the same cool-wet Monte Arido highlands
mentioned earlier. These two streams and their tributaries are
thought to con-tain most of the steelhead habitat in the Santa
Clara system, though two other creeks, Santa Paula and Hopper, also
contain significant steel-head habitat. Moore (1980a), Kelley
(2004) and Kelley and Stoecker (2005) provide overviews of Santa
Clara steelhead, and Dvorsky (2000) de-scribes a focused study of
geomorphic influences on O. mykiss distribution in Sespe Creek.
Finally, at the southern end of the NOLA sec-tion are the Santa
Monica Mountains, drained by a series of small south-facing basins
somewhat like those on the Santa Barbara coast, but drier (Plate
V). Of these basins, Malibu Creek is similar to Gaviota Creek in
that it penetrates the mountain range and drains a portion of its
north slope. The rest of the north slope is drained by Calleguas
Creek, which runs west to the ocean; and the headwaters of the Los
Angeles River, which runs around the eastern flank of the mountain
range. The southern-most steelhead creek in the NOLA section is
Topanga Creek, which harbors a small steelhead population (Dagit
and Webb 2002).
2.3. SOLA Section At the town of Santa Monica, the coast
departs
from the feet of the mountains. More than half of the large,
thick alluvial fan now inhabited by 20 million people was deposited
in recent geological times (Gumprecht 1999), from sediments washing
out of the rapidly growing mountains to the north.
For steelhead, this means that there are no mountain streams
close enough to the coast to benefit in summer from the ocean’s
climatic cool-ing effect. The tall mountains apparently have cool
temperatures suitable for steelhead and trout, but are further
inland, at the far ends of the Santa Ana and San Gabriel River
systems (Plate III, Plate VI).
The aridity of the region probably hinders the migration of
steelhead up and down the main-stems. For example, the
southern-most of the cool-wet areas is the San Jacinto mountain
range south-west of Palm Springs. Its western faces are drained by
the San Jacinto River, which theoretically
drains to the Pacific Ocean via Lake Elsinore, Te-mescal Wash
and thence the Santa Ana mainstem. In fact it does so only in very
wet years (Figure 6).
The wettest spot in the entire SOLA study area, for the period
1961 – 1990, is the northwest part of the Santa Ana basin,
specifically the Cuca-monga Wilderness west of Cajon (clearly
visible in Plate VI). During the 20th Century, discharge in the
principle creeks appears to have been more reli-able here than in
the San Jacinto River mentioned above (Figure 9). Yet Figure 9
clearly shows that many years had virtually no discharge and hence
few migration opportunities.
South of the Santa Ana basin are a series of elongated basins
draining Orange and western San Diego Counties. Climate maps
suggest that August air temperatures here are typically at least
20° C (Plate III), so the maintenance of cool stream temperature,
where it occurs, seems likely to de-pend on non-climatic factors,
such as inputs of groundwater. There is, remarkably enough, a
well-documented steelhead population in one of the smaller of these
coastal basins, San Mateo Creek (Hovey 2004). Estimated size of the
breed-ing population (never accurately determined) was thought to
be less than 70 individuals by Hovey (2004).
Figure 6. Days per year in which mean discharge ex-ceeded 30 cfs
under a natural flow regime, for the po-tential migration corridor
draining the San Jacinto Mountains. The period illustrated is prior
to the use of the San Jacinto River as an aqueduct, initiated in
1956.
-
15
Figure 7. NOLA Section, showing principal streams, towns, and
mountain ranges.
-
16
Figure 8. SOLA Section, showing principal streams, towns, and
mountain ranges.
-
17
The original distribution of O. mykiss in the SOLA section,
prior to European colonization, is not well known; in this report
we have summa-rized in the appendix (§10.3, p. 101) some of the
historic record that is known to us.
2.4. A Working Definition of Population
On page 3 we described viable, independent populations as the
primary components of a re-covery plan, where “viable” and
“independent” were defined as in McElhany et al. (2000). How-ever,
before we apply these concepts we need a working definition of the
term “population” itself. We adopt the following convention:
A population is a group of fish and their progeny that share a
reasonable expectation of interbreeding, judged by their likelihood
of co-occurrence in a stream segment during the winter migration
season.
In the study area most of the coastal basins are small enough
that one could reasonably expect all the fish inhabiting a
particular basin to constitute a single population. In addition to
this there are en-vironmental forces encouraging the fish to move
around the basin and commingle. For example,
Payne and Associates (2001) conducted an exten-sive study of
juvenile distribution in the San Luis Obispo Creek system during
summer 2001. Their data suggest that the juveniles from all over
the watershed tend to congregate in the lower part of the mainstem
creek during the summer because this reach has reliable discharge.
Many of the tributaries of the mainstem dry up during the summer
(e.g. Figure 3 in Payne et al. 2004), and in dry years may not
support breeders during the winter. This pattern appears to us to
be typical for the study area, and may sometimes force the spa-tial
co-occurrence of breeders originating from different tributaries.
This suggests that as a gen-eral rule all the O. mykiss in a
coastal basin should be grouped into a single population.
Can we expect fish in different coastal basins to interbreed?
This would require either juvenile movement through the
ocean—believed to be ex-tremely rare—or adult dispersal.
When steelhead return to freshwater to spawn, they occasionally
stray to the mouths of non-natal systems, a phenomenon known as
dis-persal. However, biologists generally believe dis-persal to be
a somewhat rare event. The historical basis for this belief in
coastal California is a study by Shapovalov and Taft (1954), who in
the 1950s
Figure 9. Discharge patterns of two creeks draining the
Cucamonga Wilderness, both tributaries of the Santa Ana River. Many
years appear to have insufficient flow for migratory access by
steelhead.
-
18
studied the steelhead populations in Scott and Waddell Creeks,
two small neighboring coastal systems in northern Santa Cruz County
(mouths 7 km apart). In an intensive tagging study, they
de-termined that only about 2-3% of a run ascended the neighboring
stream rather than their natal stream. The general assumption today
appears to be that this figure may be biased high—because streams
whose mouths are further apart on the coast are assumed to exchange
even smaller per-centages; and because Scott Creek had a hatchery
and hatchery fish are believed to exhibit higher–than-natural
dispersal rates.
Rarity of dispersal is corroborated by recent genetic studies.
Garza et al. (2004) describe a ge-netic tree of relatedness for
steelhead from 41 ba-sins throughout coastal California. They found
a pattern of isolation-by-distance among the basins, and also that
the terminal branch lengths of the tree tended to be much longer
(and better sup-ported) than the internal branch lengths. This last
result suggests that each basin has a fairly distinct genetic
population, and has relatively small amounts of genetic exchange
with neighboring basins.
Based on these studies, a “one basin = one population” rule is a
reasonable working hypothe-sis.
However, dispersal rate may vary geographi-cally due to local
adaptation, and this could cause much movement between individual
coastal ba-sins under some circumstances. Theoretical work on the
evolution of dispersal suggests that high dispersal is most likely
to evolve when the bene-fits of not dispersing are unreliable
(Johnson and Gaines 1990). This is a definite possibility in the
study area—stream discharge and thus migration access appears to be
less reliable in the study area than in northern California or the
Pacific North-west; and this would tend to select for an
oppor-tunistic flexibility in the homing tendencies of salmonids.
No such tendency has been demon-strated for the steelhead in the
study area. One piece of information that suggests such a tendency
is the rapid return of steelhead to the Carmel River after the
river was “re-watered” in the mid 1990s. However, these data could
also be interpreted as regeneration of the anadromous form of the
fish
by freshwater residents in the headwater tributar-ies (which did
not dry up).
Streams in the study area typically have sand-bar barriers at
their mouths during the dry season, transforming the estuary to a
freshwater lagoon. In years with low rainfall, these barriers are
com-monly observed to persist throughout the rainy season as well,
and this too suggests that migra-tion access is unreliable and
forces the steelhead to be flexible and opportunistic in their
migration behavior.
If the steelhead in the study area have unusu-ally high and
opportunistic dispersal patterns, it might tend to knit the
steelhead of multiple basins together into a single population—in
other words, an exception to the one-basin-one-population rule. The
situations in which this scenario seems most plausible are 1) sets
of small neighboring basins, such as in Big Sur, the southern Santa
Barbara coast, and the Santa Monica Mountains; and 2) neighboring
basins with unreliable flow, such as those in the SOLA section of
the study area.
There is also the possibility that some of the larger basins may
contain more than one popula-tion. This is especially likely for
the very large Salinas Basin, and in §2.6 and §10.1 we examine this
question more carefully.
For recovery-planning overall we suggest a prudent approach: In
the short term, adopt the one-basin-one-population rule as a
default. But, for particular basins in which a compelling argu-ment
suggests an alternative population structure, assume the
alternative structure. Over the longer term, it would be useful to
conduct research on the movement patterns of steelhead,
particularly in the Big Sur, Santa Barbara Coast, and in the
steel-head-inhabited parts of the SOLA section of the study
area.
2.5. Historic Steelhead Populations Given the
“one-basin-one-population” rule, it
is straightforward to make an accounting of origi-nal
populations using historical accounts from Titus et al. (2003),
Franklin (1999), Stoecker and CCP (2002), and Sleeper ( 2002).
These accounts provide evidence for occurrence in 87 of the 173
-
19
basins in the study area(listed in Table 2 on p. 21). Two points
are worth bearing in mind.
The first point is that the list of original popu-lations in
Table 2 are mostly based on observations of juvenile O. mykiss in
so-called “anadromous waters” at some point during the 20th Century
(Ti-tus et al. 2003). Anadromous waters are reaches believed to be
accessible to fish swimming up-stream from the ocean during their
migration sea-son (Jan. – May). However, O. mykiss in such reaches
may not necessarily be expressing the anadromous life-form at a
given time—they may be freshwater residents.
The second point is that absence from the list means absence of
evidence for fish; not necessarily absence of the fish themselves.
There had been no recent systematic attempt to locate steelhead in
all of the 173 coastal basins until 2002, at which point Boughton
et al. (2005) managed to survey 132 of them. In the process they
discovered O. mykiss in four coastal basins not mentioned in Titus
et al. (2003) or other sources. The newly-documented steelhead
basins were Malpaso Creek, Vicente Creek, and Villa Creek in
Monterey County (Boughton et al. 2005); and Los Osos Creek in San
Luis Obispo County (Payne and Associates 2001).
In some basins, Boughton et al. (2005) did not observe O.
mykiss, and if the basin had a historical account of the fish they
classified the population as extirpated or as excluded from their
habitat by anthropogenic barriers (Table 2). The extirpation
classification was based on spot checks of the best-occurring
summer habitat in the basin. “Best-occurring” was a subjective
designation stemming from field reconnaissance by an experienced
team of researchers; the subsequent spot-check con-sisted of a
snorkel survey along a 100m transect. Naturally, this
rapid-assessment technique may miss some extant populations. Still,
Boughton et al. (2005) found that if juvenile steelhead were found
at all, they tended to be observed within the first 30m of the
survey transect. This suggests that ju-venile populations tend to
fall into two categories: dense enough to be easily detected in a
100m tran-sect, or completely absent. In 17 cases Boughton et al.
(2005) were able to conduct replicate spot-checks in different
parts of a basin, always finding the same result as the initial
spot-check. Conse-quently, Boughton et al. (2005) suggested the
rapid-assessment technique probably had a rea-sonably low error
rate, and most of the apparent extirpations were true
extirpations.
Figure 10. The mouth of the Salinas River at Elkhorn Slough in
1854. Elkhorn Slough is labeled “Es-tero Grande or Roadhouse
Slough” on the map.
-
20
Even so, a more intensive study might turn up additional extant
populations, either because error rates were higher than thought,
or because some of the vacant basins were subsequently colonized.
For example, the latter may have occurred in the San Juan system in
Orange County. Boughton et al. (2005) conducted four separate spot
checks in this basin in 2002 and reported no evidence of ju-venile
steelhead; since that time several people have reported sporadic
adult migrants1. There are no reports so far of successful
reproduction (i.e., juveniles).
Some of the migration barriers reported by Boughton et al.
(2005) may turn out, after more intensive study, to be better
described as migra-tion impediments. Since many of these streams
have extant populations of O. mykiss above these impediments, these
populations might justifiably be reclassified from
freshwater-resident to possi-bly anadromous. Unfortunately, the
passability of small instream barriers by adult steelhead appears
to be an intricate and poorly understood subject. Opinion varies
widely about the abilities of steel-head with respect to barriers
and impediments.
In addition to the historical steelhead basins listed in Table
2, we also list so-called non-historical basins in Table 3. These
are basins for
1 M. Larson, personal communication, CDFG.
which no one has yet described observations of O. mykiss,
according to Boughton et al. (2005), Titus et al. (2003), Stoecker
& CCP (2002), Sleeper (2002), and Franklin (1999). Ed Henke, of
Ashland Ore-gon, has reportedly compiled historical accounts for
some of these basins but has not yet made them public.
One basin in Table 3 deserves special mention. Elkhorn Slough,
listed as one of the 173 coastal basins, could reasonably be viewed
as a part of the Salinas River system. When first mapped in the
19th Century the current northwest arm of the slough was actually
the mouth of the river, and the slough proper was a side-bay on
river-right (Figure 10). At that time the slough proper was
relatively shallow at low tide (deepest: 1.5m; Van Dyke et al.
2005), and might have served as an im-portant steelhead rearing
area. During the past millennium, the mouth of the Salinas has
probably alternated repeatedly between its 1854 location and its
present-day location 8 km south (Gordon 1996).
-
21
Table 2. Coastal basins historically occupied by steelhead, with
data on recent occupancy in anadro-mous waters1
S.-Central California Coast Steelhead ESU Southern California
Coast Steelhead ESUCoastal Basin (N to S) Extant?3 Coastal Basin (N
to S) Extant?3
Pajaro River Y Santa Maria River Y Salinas River Y Santa Ynez
River Y Carmel River Y Jalama Creek Negative obs. San Jose Creek Y
Canada de Santa Anita Y Malpaso Creek2 Y Canada de la Gaviota Y
Garrapata Creek Y Canada San Onofre Negative obs. Rocky Creek Y
Arroyo Hondo Y Bixby Creek Y Arroyo Quemado Barrier Little Sur
River Y Tajiguas Creek Barrier Big Sur River Y Canada del Refugio
Negative obs. Partington Creek Y Canada del Venadito Barrier Big
Creek Y Canada del Corral Barrier Vicente Creek2 Y Canada del
Capitan Negative obs. Limekiln Creek Y Gato Canyon Not determined
Mill Creek Y Dos Pueblos Canyon Barrier Prewitt Creek Y Eagle
Canyon Not determined Plaskett Creek Y Tecolote Canyon Barrier
Willow Creek - Monterey Y Bell Canyon Barrier Alder Creek Y Goleta
Slough Complex Y Villa Creek – Monterey2 Y Arroyo Burro Barrier
Salmon Creek Y Mission Creek Y San Carpoforo Creek Y Montecito
Creek Y Arroyo de la Cruz Not determined Oak Creek Barrier Little
Pico Creek Not determined San Ysidro Creek Y Pico Creek Not
determined Romero Creek Y San Simeon Creek Y Arroyo Paredon Y Santa
Rosa Creek Y Carpinteria Salt Marsh Complex Barrier Villa Creek -
SLO Y Carpinteria Creek Not determined Cayucos Creek Negative obs.
Rincon Creek Barrier Old Creek Dry Ventura River Y Toro Creek Y
Santa Clara River Y Morro Creek Y Big Sycamore Canyon Negative obs.
Chorro Creek Y Arroyo Sequit Y Los Osos Creek2 Y Malibu Creek Y
Islay Creek Y Topanga Canyon Y Coon Creek Y Los Angeles River
Barrier Diablo Canyon Y San Gabriel River Barrier San Luis Obispo
Creek Y Santa Ana River Barrier Pismo Creek Y San Juan Creek
Negative obs. Arroyo Grande Creek Y San Mateo Creek Y San Onofre
Creek Dry Santa Margarita River Negative obs. San Luis Rey River
Barrier? San Diego River Barrier Sweetwater River Barrier Otay
River Barrier Tijuana River Not determined 1 Historical data: Titus
et al. (2003); Sleeper (2002); Franklin (1999). Recent data:
Boughton et al. (2005) 2 No data on historical occurrence, but
recent occurrence documented by Boughton et al. (2005). 3 “Negative
obs.” means juveniles were observed to be absent during a
spot-check of best-occurring summer habitat in 2002. “Dry”
indicates the stream had no discharge in anadromous reaches during
the summer of 2002. “Barrier” indicates that all over-summering
habitat was determined to be above an anthropogenic barrier,
believed to be im-passable. See Boughton et al. (2005) and notes on
page 19 for details.
-
22
Table 3. Coastal basins with no evidence1 of historical or
extant populations of O. mykiss in anadro-mous waters.
South-Central California Coast Steelhead ESU
Southern California Coast Steelhead ESU
Coastal Basins (N to S) Coastal Basins (N to S) Elkhorn Slough
Shuman Canyon Santa Monica Canyon Del Rey Creek San Antonio Creek
Solstice Canyon Seal Rock Creek Honda Creek Corral Canyon Pescadero
Canyon Creek Wood Canyon Trancas Canyon Gibson Creek Damsite Canyon
Escondido Canyon Soberanes Creek Canada del Cojo Ramirez Canyon
Doud Creek Barranca Honda Zuma Canyon Sycamore Canyon Canada de la
Llegua Ballona Creek Grimes Canyon Arroyo San Augustin Dominquez
Channel Hot Springs Canyon Arroyo El Bulito San Diego Creek Lime
Creek Canada del Agua Los Trancos Canyon Kirk Creek Canada de la
Cuarta Muddy Canyon Wild Cattle Creek Canada de Alegria Moro Canyon
Soda Spring Creek Agua Caliente Emerald Canyon Arroyo de los Chinos
Canada del Molino Laguna Canyon Oak Knoll Creek Las Llagas Canyon
Aliso Creek Little Cayucos Creek Las Varas Canyon Salt Creek Willow
Creek - SLO Sycamore Creek Canada de Segunda Deshecha Little Irish
Canyon Toro Canyon Creek Las Pulgas Canyon Irish Canyon Los Sauces
Canyon Aliso Canyon Wild Cherry Canyon Hall Canyon Loma Alta Creek
Pecho Creek Arundell Barranca Buena Vista Creek Rattlesnake Canyon
Calleguas Creek Agua Hedionda Creek La Jolla Canyon Canyon de las
Encincas Little Sycamore Canyon San Marcos Creek Carbon Canyon
Escondido Creek Las Flores Canyon San Dieguito River Piedra Gorda
Canyon Los Penasquitos Creek Pena Canyon Rose Canyon Tuna Canyon
Tecolote Creek Santa Ynez Canyon Chollas Creek Telegraph Canyon 1
No evidence of occurrence does not imply evidence of no occurrence;
the latter would require a “nega-
tive observation;” i.e. evidence that the species was looked for
but not found.
-
23
2.6. Three Discrete Populations in the Salinas System
The Salinas Basin appears to possess five dis-tinct steelhead
areas—Gabilan Creek, Arroyo Seco, the San Antonio River, the
Nacimiento River, and the upper Salinas system south of San Miguel.
The large size of the basin suggests that these steel-head areas
may be sufficiently isolated to contain multiple populations—a
possible exception to the one-basin/one-population assumption. A
detailed assessment of this hypothesis is in §10.1 (p. 92 of the
Appendix); here we summarize the main con-clusions.
Gabilan Creek should be considered a distinct population, due to
its unique connection with the ocean via Tembladero Slough and the
Old Salinas River channel which connects Elkhorn Slough to the
current estuary of the Salinas River. The steel-head habitat is
high in the headwaters of the sys-tem, quite isolated from the
other steelhead areas by low-gradient sloughs in the lower Salinas
Val-ley.
Arroyo Seco should probably be considered a distinct population,
for three reasons. First, it is separated from the other steelhead
areas by a long stretch of the mainstem Salinas that does not
ap-pear particularly hospitable to juvenile movement. This appears
to be an isolating mechanism. Sec-ond, there is no evidence that
under natural hy-drologic conditions, low streamflow prevented
adults from homing to Arroyo Seco (This would potentially force
them to spawn elsewhere in the basin, presumably in the other
steelhead areas). Third, the consequences of making a “lumping”
mistake (erroneously treating Arroyo Seco as not distinct) appear
to be greater than making a “split-ting” mistake (erroneously
treating it as distinct).
Nacimiento, San Antonio, and upper Salinas Rivers should be
considered to jointly share a sin-gle distinct population. There is
evidence that un-der natural hydrologic conditions, low streamflow
often prevents adult migrants from returning to a particular
stream, forcing them to spawn in one of the other two steelhead
areas comprising the population. Under natural hydrologic
conditions, the discharge from the Nacimiento River appears to be
the most reliable, so usually it would be fish natal to the San
Antonio or upper Salinas that would be forced by low flows to spawn
in the Nacimiento. Thus we refer to this population as “Nacimiento
et al.” This three-population scheme for the Salinas Basin is based
on indirect evidence pertaining to steelhead movement patterns.
Data on fish movement between the five steelhead areas would give a
clearer and more accurate picture of the population structure of
the Salinas Basin. More detail on the indirect evidence and its
evaluation is given in the appendix, §10.1 (p. 92). At present the
mainstem of the Salinas River does not appear to comprise suitable
spawning or rearing habitat for steelhead, and this is partially
the basis for considering the basin to have several discrete
populations in major sub-drainages. However, the mainstem prior to
Spanish settle-ment may have been quite different ecologically,
having a well developed riparian forest; a higher water table
providing cool water inflows to the mainstem; a less incised
channel; and substrates coarser than the sand and silt that now
predomi-nate. These conditions would have been more conducive to
steelhead spawning and rearing than the current state of the river.
The evidence that the mainstem was once more suitable for steelhead
is discussed in the Appendix (§10.2, p. 98).
-
24
Part 3. Extant Populations
In 2001, Boughton et al. (2005) undertook a survey of the
original populations listed in Table 2 (p. 21), to determine which
had extant anadro-mous components. In 2002 a similar survey was
conducted in the sub-basins of the two largest sys-tems, the
Salinas and Pajaro. Overall, 92 historical steelhead basins and
sub-basins were identified, and occurrence was estimated in 86 of
them through a combination of literature review, field
re-connaissance, and snorkel surveys (spot checks) in the
best-occurring habitat. Six basins were inac-cessible for various
reasons and were not sur-veyed. The survey also assessed occurrence
in 55 coastal basins that lacked any historical account of
steelhead occurrence, and detected the species in 5 of them.
The survey indicated that between 58% and 65% of the historical
steelhead basins currently harbor O. mykiss populations at sites
with migra-tion connectivity to the ocean. Most of the appar-ent
losses of steelhead occurred in the south (Figure 11).
Sixty-eight percent of the apparent basin-wide losses of
steelhead were associated with anthro-pogenic barriers to fish
migration (dams, flood-control structures, culverts, etc.; termed
“barrier exclusions” in Figure 11). According to a regres-sion
analysis, the barrier exclusions were statisti-cally associated
with highly-developed water-sheds (percent cover by urban and
agricultural development). The remaining losses of steelhead (not
associated with migration barriers) had no statistical association
with percent development in the watershed; they tended to occur in
basins with relatively warm climate (mean annual air tem-perature
as inferred using the parameter-elevation regressions on
independent slopes model; Daly et al. 1994).
The lowest rate of basin-wide steelhead loss (zero) was along
the Big Sur coast. Every one of the 21 original populations between
the towns of Carmel and Cambria had extant O. mykiss in ana-dromous
waters, and 3 additional populations were also detected by Boughton
et al. (2005).
Nearly all of these occupied basins are extremely small in area,
but are located in one of the wettest and coolest parts of the
study area. The region is sparsely settled, and the steelhead
populations do not appear to be heavily impacted by water
diver-sions, habitat alteration, etc.
In general, the apparent losses or extirpations were inferred
from spot checks and involve a cer-tain amount of error. At the
level of individual spot checks, the estimated probability of
detection failure was 0.0175 (Boughton et al. 2005). At the basin
level, error was estimated from 17 revisits to novel sites in
basins already spot-checked, and in all cases the result (detected;
not detected) was the same as that of the previous spot check,
giving an error probability of 0.00 (95% c.i.: [0.00, 0.162] by the
binomial distribution). In addition, the basins classified as
“barrier exclusions